The story
She's mid-shoot. Four stylist hands fuss at her curls and lapels. She raises a palm and stops the take.
What follows is an actor explaining, with total sincerity, how a render trigger should be performed — a slow press for a slow render, traced with an imaginary tear down the cheek; an intense render, clawed at the air. The director, tired, points out it's just a render trigger.
"A boring render is a form of trauma."
She says trust me, squares her shoulders — and the film hard-cuts to black. Cool light, specular edge, the trigger held below her lips, dead-eyed:
"To generate, or not to generate... that was never the question."
Product beauty shot. Cut back to the warm set. The director says "perfect", and she collapses into real laughter.
What's The Generation Sequence for This?
| Step | Where | Model |
|---|---|---|
| 1–3 | Image | Nano Banana Pro — the warm set, the trigger, the cool hero frame |
| 4–13 | Video | Seedance 2.5 I2V — fourteen generations |
| 14 | Audio | ElevenLabs — dialogue, then the two sound worlds |
Before you start — three settings that decide everything
Seedance 2.5 I2V has no aspect-ratio parameter. Output follows the source image, so every still you make must be 9:16. Frame the picture the way you want the video framed.
Set generate_audio off. You want the dialogue performance built
separately — see step 14. Generated speech fights the lip sync you'll add later.
Two visual worlds, two reference families. The warm set and the cool hero sequence never blend. Don't try to morph between them inside a generation; the hard cut is the joke.
Step 1 — The warm set
MODEL: GenX Pro -> Image → Nano Banana Pro.
Settings: 9:16.
Use this still as the I2V input for warm-set shots.
Vertical 9:16 medium portrait on a premium beauty commercial set. A woman in her
late twenties sits centred on a stool, framed from mid-chest up. Warm brown
skin, natural dark curly hair worn full and shoulder-length, small gold hoop
earrings, clean editorial makeup with a soft matte finish. Calm, composed,
neutral expression, looking just off camera.
Wardrobe: matte black fine-knit turtleneck under a sharply tailored black
blazer, no buttons fastened, no lapel pin, no visible labels.
Background: a deep terracotta painted canvas backdrop, softly mottled, filling
the frame behind her with no visible edges or seams.
Lighting: a single large diffused key from camera-right at head height, deep
soft shadow falling across the far side of her face, gentle falloff into the
backdrop, rich warm skin tones with visible pore detail, catch-lights in both
eyes. Shot on a 85mm lens at f/2, shallow depth of field, the backdrop softly
out of focus. Editorial commercial photography, fine natural grain.
Clean frame, textless, unbranded. No text, no logos, no props.
The terracotta is doing structural work. It's warm, expensive and slightly theatrical — it makes the set read as a real commercial shoot, which is what sells the later cut to black. A grey or white backdrop kills it.
Say "clean frame, textless, unbranded" on every prompt in this build. Both image and video. It's the phrase that stops invented lettering appearing on the backdrop, the wardrobe and the product.
Step 2 — The trigger
MODEL: GenX Pro -> Image → Nano Banana Pro.
Settings: 9:16.
Vertical 9:16 premium product photograph of a sleek handheld device standing
upright, centred, against a deep charcoal-black background. The device is a
narrow rounded matte black body, roughly the size of a TV remote, with a single
circular button set into the flat top, ringed by a thin precise band of white
illuminated light.
Cool specular light from top and side traces the silhouette and picks out the
soft sheen of the matte finish. Faint atmospheric haze in the background.
Controlled reflections, no blown highlights, the white ring perfectly circular
and evenly lit. Razor-sharp product detail, very shallow depth of field.
Clean frame, textless, unbranded. No hands, no text, no logo, no markings of
any kind on the body.
Give the prop one memorable feature and hold it. The glowing white ring is the whole silhouette — it's what makes the macro shots read and what tells you instantly when a generation has drifted.
Step 3 — The cool hero frame
MODEL: GenX Pro -> Image → Nano Banana Pro.
Settings: 9:16.
Vertical 9:16 extreme close-up beauty portrait against a rich black background.
The same woman — warm brown skin, dark curly hair, small gold hoop earrings —
her face filling the frame, direct piercing eye contact into the lens, lips
closed, expression severe and still.
She holds a matte black handheld device vertically just below her lips, its
circular top button glowing with a thin white ring, the device sharp in the
foreground.
Lighting: hard specular rim light from behind and above separating her from the
black, cool neutral key, minimal beauty diffusion, skin texture crisp and
controlled rather than soft. Faint cool particulate drifting in the depth
behind her. Desaturated, cool, deep blacks with no lift.
Luxury product-film realism. Clean frame, textless, unbranded, no typography,
no logo.
Ask for less beauty diffusion here, not more. The instinct is to make the hero shot glossier. It should be harder — crisp skin texture and a cold rim light is what makes it read as the parody of a fragrance ad rather than an actual one.
Steps 4–13 — The film
MODEL: GenX Pro -> Video → Seedance 2.5 I2V.
Settings: 720p · generate_audio off.
Fourteen generations. Every one starts from a still — either one of your three references, or the last frame of the previous generation.
Chaining. Where a shot says "continue from the last frame", export the final clean frame of the previous generation and feed it in as the source image. Cut on motion or on a blink; a 2–4 frame overlap hides a small pose mismatch.
| Beat | Length | Generations | |
|---|---|---|---|
| 1 | Slate | 0.7s | 1 |
| 2 | The interruption | 4.6s | 1 |
| 3 | Director over-the-shoulder | 1.5s | 1 |
| 4 | Trigger macro | 3.6s | 1 |
| 5 | Emotional extremes | 8.7s | 2, chained |
| 6 | Reality check | 9.2s | 2, chained |
| 7 | Trust me + reset | 6.5s | 2, chained |
| 8 | Hero sequence | 8.8s | 3 |
| 9 | Product beauty shot | 3.0s | 1 |
| 10 | The break | 5.6s | 1 |
1 — Slate · 1s
Extreme close-up of a sterile white film clapperboard filling the vertical
frame. A hand snaps the black-and-white top bar shut once, fast and decisive.
Neutral production lighting, shallow depth of field, clean white surface,
realistic live-action motion, locked camera, no readable branding, textless,
unbranded.
2 — The interruption · 5s
Source image: the warm set still.
Vertical medium portrait on a premium beauty set. Actor sits centred in a matte
black turtleneck and tailored black blazer against a deep terracotta backdrop.
Four stylist hands enter from both sides adjusting curls, earrings and lapels.
She begins stoic, then suddenly raises one palm to stop the take and gives a
controlled, slightly annoyed interruption. Diffused key from camera-right, deep
soft shadows, shallow depth, locked camera, clean frame, textless, unbranded.
3 — Director over-the-shoulder · 1.5s
Soft-focus over-the-shoulder view from behind a seated director wearing a dark
cap. His head and shoulder dominate the lower foreground, a metal cinema camera
rig crosses the centre, and the actor in black sits sharp against the terracotta
backdrop beyond. The director makes a tiny tired head movement as if sighing.
Actor waits, still. Warm studio light, shallow depth, premium live-action set,
clean frame, textless, unbranded.
4 — Trigger macro · 3.6s
Source image: the trigger still.
Extreme macro close-up of a sleek matte black handheld render trigger. A
realistic thumb hovers a few millimetres above a circular top button with a thin
glowing white ring. The thumb hesitates and subtly trembles but never presses.
Warm skin tones, dark background, one large soft golden bokeh orb behind,
razor-sharp product and fingertip, very shallow focus, controlled reflections,
clean frame, textless, unbranded.
It never presses. The whole film is deferral — she talks about the render instead of rendering. The hovering thumb is that idea in one shot.
5A — The fake tear · 4.3s
Locked frontal medium shot on the terracotta beauty set. Actor in black holds
the small trigger in one hand. With the other index finger she traces one
imaginary tear down her cheek, demonstrating an absurdly emotional slow render.
Expression shifts from solemn concentration to delicate melodrama while her
torso stays composed. Camera rig remains low in foreground. Warm diffused key,
deep soft contrast, shallow depth, clean frame, textless, unbranded.
Export the last clean frame.
5B — The claw · 4.3s
Source image: last frame of 5A.
Continue seamlessly from the supplied last frame. Same actor, black suit,
terracotta set, framing and light. Her melodramatic sadness snaps into an
intense render demonstration: eyes widen, jaw tightens, and her free hand claws
the air toward camera in one strong controlled gesture while the trigger stays
secure in the other hand. Keep body position and camera rig unchanged. Premium
beauty lighting, shallow depth, clean frame, textless, unbranded.
Cut 5A to 5B on the start of the hand acceleration.
6A — She absorbs it · 4.6s
Warm terracotta beauty set, tight medium portrait of the actor in the same matte
black turtleneck and blazer. She looks toward the off-camera director, holding
the black trigger low. She listens to a deadpan correction, blinks once, then
gives a tiny offended inhale as if the remark is unbelievable. Keep movement
subtle and expensive, no camera move, diffused side key, deep soft shadow,
shallow depth, clean frame, textless, unbranded.
Let the joke breathe. The blink and the inhale are the entire performance. Anything bigger and the line that follows has nowhere to land.
6B — The trauma line · 4.6s
Source image: last frame of 6A.
Continue seamlessly from the supplied last frame. Same actor, set, wardrobe,
framing and light. She places one hand flat over her chest, leans back a
fraction, and responds with solemn wounded conviction as if defending a deeply
serious principle. Her brows lift slightly and her eyes stay locked toward the
director off camera. Controlled natural mouth motion for later lip sync, no
exaggerated gestures, clean frame, textless, unbranded.
Ask for controlled mouth motion, not speech. You're adding the real performance in step 14, and a lip-sync pass fights extreme generated jaw shapes. Keep the mouth neutral enough to be replaced.
7A — Trust me · 5s
Locked warm-set medium shot. Actor in black sits centred against terracotta,
camera rig low foreground. She recovers instantly, points one index finger
toward the off-camera director with calm certainty, and gives a small confident
nod as if saying trust me. Her other hand keeps the trigger. The comedy is dry,
not broad. Premium diffused side light, shallow depth, stable face and wardrobe,
clean frame, textless, unbranded.
7B — The reset · 2s
Source image: last frame of 7A.
Continue from the supplied last frame for a short reset. Same actor and warm
beauty set. She lowers the pointing hand, centres the trigger, straightens her
shoulders, lifts her chin by a few millimetres and settles into a perfectly
still luxury-ad pose with direct, serious eyes. No camera movement. End on a
clean neutral hero posture, textless, unbranded.
End clean, then hard cut to black. Don't attempt a transition into the cool world. The sudden change is the joke, and a hard cut looks more expensive than anything a model will invent.
8A — The ring emerges · 2.4s
Black cinematic void with faint cool particulate drifting in depth. A small
luminous white ring slowly emerges from darkness near the lower centre, like the
illuminated button of a premium device. Minimal movement, glossy controlled
highlight, cool desaturated atmosphere, specular top light, deep blacks, no
actor yet, no typography, no logo, clean frame.
8B — Hero close-up, first half · 3.2s
Source image: the cool hero still.
Extreme close-up beauty portrait in a black cool-toned sci-fi studio. Same
actor, face razor sharp, direct piercing eye contact. She holds the matte black
trigger vertically just below her lips, its circular button glowing with a thin
white ring. Almost no head movement, subtle controlled mouth motion and breath,
faint particles behind, hard specular rim and top light, rich black background,
luxury product-film realism, no typography, no logo.
8C — Hero close-up, second half · 3.2s
Source image: last frame of 8B.
Continue seamlessly from the supplied last frame. Same extreme close-up, black
cool sci-fi studio, trigger position, light and direct eye contact. She finishes
the thought with tiny controlled lip and eyebrow movement, then becomes
perfectly still for the final beat. Keep the glowing ring stable and circular.
No head turn, no camera move, no new objects, luxury product-film realism, no
typography, no logo, clean frame.
Identity stability beats drama here. Almost nothing moves in 8B and 8C, and that stillness is what makes the shot read as a real fragrance ad. Every extra motion you ask for is another chance for the face to drift.
9 — Product beauty shot · 3s
Source image: the trigger still.
Premium vertical product beauty shot of a sleek matte black handheld trigger
floating or standing centred against a deep charcoal-black background. Narrow
rounded body, circular top button with a precise thin white illuminated ring.
Cool specular top and side light traces the silhouette, soft atmospheric haze,
subtle depth, slow elegant micro push-in or slight product drift, controlled
reflections, no hands, no typography, no logo, clean frame.
10 — The break · 5s, retimed to 5.6s
Back on the warm terracotta beauty set, locked medium close-up. Same actor in
black begins perfectly serious, holding the trigger toward camera. She hears an
off-camera director approve the take; after a tiny delay her intense persona
collapses. She smiles, bursts into genuine spontaneous laughter, shoulders
loosen, and her head dips and turns partly out of frame. Keep the laugh natural
and joyful, face stable, warm diffused side light, textless, unbranded.
Generate five seconds and retime to about 90% speed in the edit. A high-energy facial performance is safer stretched than chained.
Step 14 — Sound
Picture first, performance second, mix last. Cut the film silent and to length before recording anything — that fixes exactly how much dialogue time each shot has.
MODEL: GenX Pro -> Audio → ElevenLabs.
Record the lines below. Use a dry, intimate voice for the actor and a roomier production-set sound for the director.
| Shot | Line |
|---|---|
| 2 | (off camera) "And action!" — then her interruption |
| 3 | Director, tired: "What is it now?" |
| 5 | Her, explaining the slow press versus the intense render |
| 6 | Director: it's just a render trigger. Her: "A boring render is a form of trauma." |
| 7 | Her: "Trust me." |
| 8 | Her, intimate whisper: "To generate, or not to generate... that was never the question." |
| 10 | Director, off camera: "Perfect!" |
Then lip-sync only the dialogue-facing shots — 2, 6A, 6B, 7A, 8B and 8C. Leave the macro and product shots alone.
Build two sound worlds, and make them different. The warm set should sound roomy and imperfect: room tone, clothing and hair movement, crew rustle, the trigger being handled, the slate. At the cut to black, strip all of it away and replace it with a low sub swell, airy particles, a controlled device hum and a crisp tonal accent on the glowing ring. Restore the warm room the instant you cut back for the laugh.
That contrast sells the parody as hard as the picture does. If both halves sound the same, the hero sequence just looks like a different shot rather than a different film.
Assembly
Warm set: keep the terracotta rich, skin warm and controlled, wardrobe matte black, blacks deep without crushing the hair.
Hero sequence: drop saturation, cool the shadows, raise specular separation, and protect the white ring from clipping.
Normalise exposure shot to shot after generating — prompt language alone won't give you consistent contrast across fourteen clips. Add a light grain pass globally once both worlds are balanced, so the film reads as captured rather than assembled.
Keep both transitions hard cuts: to black at the hero sequence, back to terracotta for the laugh.
Make it your own
Log in to GenX Pro or create a free account, then open the App and try these prompts.